Full Day: 2nd International Workshop on Spatial Cognition and Artificial Intelligence (SC&AI) (Vasardani, Richter)

Click below to see details 



Maria Vasardani


Kai-Florian Richter (contact person)




The 2ndInternational Workshop on Spatial Cognition and AI(SC&AI) is organized by the Artificial Intelligence & Cognitive Science Working Group of TC12, the International Federation for Information Processing Technical CommitteeonArtificial Intelligence (AI), and open to anybody interested in the topic. This Working Group, and consequently the workshop, is born from the re-emerged need to bring together members of the international scientific community who work on both theoretical as well as applied overlapping research areas of the two fields.

We believe that in today’s era of deep learning and ‘black-box’ AI that appears in ‘spatial’ agents, such as self-driving vehicles, service robots, and interactive location-based services (among others), the interdisciplinary nature of spatial cognition and its strong connection with AI may need to become more evident than ever. It takes a joint effort to make such cognitively inspired artificial systems a success.

In this spirit, the 2nd SC&AI Workshop aims at bringing together researchers from different disciplines (e.g., AI & computer science, cognitive psychology & neuroscience, GIScience, cognitive geography & cartography, linguistics, philosophy of mind, architecture and design, engineering, mathematics, and others) to work on cognitively inspired spatial artificial systems.



This full-day workshop will address issues of shared spaces between human and artificial system, and the blurring of physical separation of human and machine in spatial tasks and interactions.
The use of mobile computing devices (smartphones, tablets, etc.) anytime and everywhere has become ubiquitous today, enabling in-situ interaction, rapid switches between physical and digital worlds, and even the blending of these worlds in augmented reality settings. But now a whole range of new spatial artificial systems seem set to enter our everyday environments, for example, computerized protheses, smart homes, social robots in households and healthcare, robotized tools in manufacture and maintenance, self-driving vehicles, and smart cities. Self-driving vehicles (and, arguably, smart homes) will for the first time make us humans interact and communicate with something that actually physically contains us! The introduction of these new systems will have effects and consequences for both system and human behavior, which we believe have not yet been sufficiently investigated so far. For example, all these systems operate on different scales and in different spatial relation to us humans (on-body, in-hand, close-by, independent-from, containing, etc.). 

We see several issues arising from this development, for both humans and systems alike. We have identified the following questions to be discussed during the workshop (which may be subject of change due to popular demand):
  • What kind(s) of (spatial) representations do such systems need in order to operate successfully? Would there be a unified representation (mechanism / framework) for all such systems?
  • How do we as human users (want to) establish common ground with such systems? How does this differ between different systems, if at all? And how might this depend on the interaction modality used?
  • What are the cognitive effects of using / interacting with such systems? Should and can we control for them?
These questions will be worked on in interdisciplinary groups of 4 to 6 participants each. The workshop will be highly interactive, and we hope to achieve some concrete outcomes. The morning sessions will be dedicated to laying the grounds by identifying concrete topics and research questions that the respective groups want to work on. In the afternoon, we will aim for concrete results and making outcomes ‘tangible’; e.g., participants may devise plans for a joint research paper (including outline, how to perform the actual research, etc.), or participants may actually decide to ‘implement’ something (similar to a hackathon).
If there is substantial interest, we will aim for a special issue in a relevant journal, offering an opportunity to continue collaboration and to get results from the workshop (and beyond) published.

Schedule for the workshop day 
(more detailed schedule including times and breaks to follow)

  • Introduction to the workshop and its aims
  • Short introduction round of all participants
  • Keynote presentation tba
  • Group challenge 1: each group will be tasked to determine which of the workshop questions they’d like to address; and then to identify a concrete scenario and research questions that could be explored
  • Presentations 1: each group will briefly outline the scenario and questions they came up with (including some reasoning as of why), followed by a discussion among all participants.
  • Group challenge 2: making outcomes of the morning tangible and opening up possibilities for concrete results, e.g.,
    • developing a joint research paper, including plans on how to do the research
    • ‘hackathon’
  • Presentations 2: Presentation and discussion of group work results: each group will have 10 minutes to present the results of tackling the challenge, followed by questions and discussion from the other participants.
  • Closing remarks and future directions, including a discussion of how the joint work started on the day may continue beyond the workshop.

Call for Participation 
If you are interested to participate, please submit an expression of interest in which you
  • briefly outline your (academic) background; 
  • discuss why you are interested in participating in the workshop (e.g., how does it relate to your work? what do you hope to get out of it?)
  • how you believe you can contribute to discussions during the day (e.g., what questions are you particularly interested in? what answers do you seek?). 

Send your expression of interest as a PDF file of not more than 3 pages by email to the workshop organizers.

Maria Vasardani, RMIT University, Melbourne, Australia
This email address is being protected from spambots. You need JavaScript enabled to view it.
Kai-Florian Richter, Umeå University, Umeå, Sweden
This email address is being protected from spambots. You need JavaScript enabled to view it.


Full Day: Speaking of Location 2019: Communicating about Space (Stock, Jones, Tenbrink)

Click below to see details






Kristin Stock (contact person)


Chris Jones


Thora Tenbrink




Research into the description of location using human (natural) language has been approached from linguistics, geospatial, and computer science perspectives.  This interdisciplinary combined workshop and tutorial will explore current developments in the area with a particular emphasis on the need for communicating about location across different contexts and for diverse purposes, and the particular challenges these differences cause for automatic generation, extraction and interpretation of natural language descriptions of geographic space. 


The scope of the event includes the following:

  • approaches to automated extraction and understanding of natural language descriptions of location in textual sources;
  • approaches to automated generation of natural language location descriptions;
  • both verbal and written geospatial natural language;
  • work on natural language related to both static and dynamic (movement) location;
  • work that aims to understand/explore/investigate the use of natural language descriptions of location;
  • research examining verbal representations of spatial problem solving and wayfinding;
  • future priorities in geospatial natural language research;
  • geospatial natural language and ontologies;
  • contextual factors in the use and interpretation of geospatial natural language;
  • real world applications that motivate the use of geospatial natural language;
  • geospatial natural language querying;
  • recent/current technological developments and their links to geospatial natural language (e.g. social media, crowdsourcing);
  • links between language and spatial cognition and behaviour;
  • robotics and geospatial natural language;
  • connections between geospatial natural language and other areas of research (e.g. qualitative spatial reasoning, semantics).



The event will include:

  • 20-minute presentations of reviewed 5-8 page papers discussing current work in progress and future research directions and ideas.Submitted papers will be peer reviewed by members of the program committee.
  • Five-minute position pitches, in which presenters give a summary of a research idea or viewpoint, culminating in the posing of a question of set of questions for discussion. These will be based on abstracts that will be checked for relevance.
  • 2-hour tutorial on Cognitive Discourse Analysis (Tenbrink, 2015): How to deal with language in spatial information and cognition research
  • A panel discussion on the workshop subtheme ‘communicating about space’, and particularly addressing the need to communicate locations across different contexts and for diverse purposes, and the challenges this places for generation and interpretation of geospatial natural language.


Cognitive Discourse Analysis Tutorial: How to deal with language in spatial information and cognition research

This short tutorial session offers practical support for the systematic analysis of language data collected in relation to studies within the scope outlined above. Typically, language data (such as location descriptions or verbal representations of spatial problems) are analysed intuitively with respect to content, addressing those aspects (e.g., particular thought processes or strategies) that the speakers are themselves aware of and that are easy to understand from what they say. However, it is often possible to gain further insights based on a closer, more systematic analysis of linguistic details.

Some aspects of language use reflect cognitive aspects that go beyond conscious reflection by individual speakers, and that are not necessarily directly observable in linguistic content – such as the perspective underlying an utterance such as ‘the chair on the right’.

The tutorial session will take the participants' current or intended projects as a starting point to address the following issues, supplemented wherever suitable by practical exercises.

Scope: What kinds of studies would benefit from a closer analysis of linguistic and conceptual features represented in the language data?

Data collection: What kinds of issues need to be considered in the light of current research purposes?

Systematic linguistic analysis: Practical aspects of systematic data annotation, substantiated by linguistic insights.

Choosing an analysis focus: Language can reveal crucial insights about cognitive aspects such as perspective, granularity, inference, certainty, and more. We will discuss aspects of interest to the audience, and explore the ways in which they relate to the participants’ projects.


Research Papers

We invite submission of papers that describe current research/work in progress, between 5 and 8 pages in length including references.  Submitted papers will be peer reviewed by members of the program committee, and accepted papers will be given a 15-20 minute presentation time slot during the workshop. Accepted papers will be published online in some form (e.g. workshop web site, University repository), and avenues for more formal publication will be discussed during the workshop.


Position Pitches

We invite submission of one-page position pitches, describing research ideas, proposals, opinions about the state of the research field or other material relevant to the workshop, and including one or more discussion questions.  Accepted position pitches will be given a 5 minute presentation time slot, followed by time for discussion of the questions posed. Submitted position pitches will be reviewed for relevance by the workshop organisers.